Week 2

We began week two by reading the two articles that Nandini instructed us to read which were, “Discriminative Boosted Bayesian Networks for Learning Multiple Cardiovascular Procedures” and “Modeling Coronary Artery Calcification Levels From Behavioral Data in a Clinical Study”.  The complexity of these two readings were initially way over my head, they discussed things such as,  Dynamic Bayesian Networks and Bayesian Dirichlet scores.  After, doing copious amounts of research and finding related articles, in particular, Kevin Murphy’s book “Machine Learning: A Probabilistic Perspective”  I was able to make a lot more sense of what a Bayesian network is, how they work, and what they are used for.  Although, I am still not completely convinced that I fully understand the intricacy of such mathematical models, I believe that I am on the right path.  From my understanding, my partner Kate, and I are going to be using such models to implement a system that can detect if a person is going to have a heart attack or not, and if they are, it will give a estimate of when the heart attack will occur.  The possibility for this to be published is still up in the air.  It all really depends on how well Kate and I can understand the models we will be using, and how well we can implement these models.

Kate and I also began to get familiar with Weka to pre-process, classify,  cluster, and visualize random data sets.  Nandini gave us a few data sets to clean, train, and then test.  Resulting in, a visualization tree that broke down our data into a organized pattern.  After this brief tutorial of Weka and gaining an understanding how to classify data as such, Dr. Natarajan instructed us to download and play around with a software called Netica.  This software enables us to create a visualization of all the aspects in a Bayesian networks and use data sets to make decision and accurate calculations that are two complex to do by hand.  This software is going to play a huge role in our project, and hopefully make our data sets a lot easier to handle.

On Friday we had to turn in a draft of our methods, related work, and proper references using ShareLaTeX.  For Kate and I, creating our methods section was a little bit more difficult than that of other groups, due to the ability for methods to constantly change while creating a machine learning program.  Lastly, Kate and I created a powerpoint slide indicating what we did throughout our second week of research at ProHealth, that we presented in our routine Friday morning meeting with the whole REU group.

 

For next week, we got assigned to read chapters 14 and 15 from the text book “Artificial Intelligence a Modern Approach”.  After we have completed the readings, Dr. Natarajan has indicated that we will be meeting with him on Tuesday for further instructions on what is to come next.

 

Our ShareLaTeX page

https://www.sharelatex.com/project/592c97104b722b286bb94c4c